對樣本資料之機率密度函數f(x)的無母數估計方法，一直是統計推論領域的研究重點之一，而且在通訊理論與圖形辨別上有非常重要的地位。傳統的文獻對密度函數的估計方法大部分著重於獨立樣本的情形。對於時間數列的相關樣本（例如：經濟指標或加權股票指數資料）比較少提到。本文針對具有弱相關性的穩定時間數列樣本，嘗試提出一個核密度估計的方法並探討其性質。 / For a sample data, the nonparametric estimation of a probability density f(x) is always one point of research problem in statistical inference and plays an important role in communication theory and pattern recognition. Traditionally, the literature dealing with density estimation when the observations are independent is extensive. Time series sample with weak dependence, (for example, an economic indicator or a stock market index data), less in this aspect of discussion. Our main purpose is concerned with the estimation of the probability density function f(x) of a stationary time series sample and discusses some properties of this kernel density.

In this thesis, we consider the problem of estimating the regression function in location-scale regression models. This model assumes that the random vector (X,Y) satisfies Y = m(X) + s(X)e, where m(.) is an unknown location function (e.g. conditional mean, median, truncated mean,...), s(.) is an unknown scale function, and e is independent of X. The response Y is subject to random right censoring, and the covariate X is completely observed. In the first part of the thesis, we assume that m(x) = E(Y|X=x) follows a polynomial model. A new estimation procedure for the unknown regression parameters is proposed, which extends the classical least squares procedure to censored data. The proposed method is inspired by the method of Buckley and James (1979), but is, unlike the latter method, a non-iterative procedure due to nonparametric preliminary estimation. The asymptotic normality of the estimators is established. Simulations are carried out for both methods and they show that the proposed estimators have usually smaller variance and smaller mean squared error than the Buckley-James estimators. For the second part, suppose that m(.)=E(Y|.) belongs to some parametric class of regression functions. A new estimation procedure for the true, unknown vector of parameters is proposed, that extends the classical least squares procedure for nonlinear regression to the case where the response is subject to censoring. The proposed technique uses new `synthetic' data points that are constructed by using a nonparametric relation between Y and X. The consistency and asymptotic normality of the proposed estimator are established, and the estimator is compared via simulations with an estimator proposed by Stute in 1999. In the third part, we study the nonparametric estimation of the regression function m(.). It is well known that the completely nonparametric estimator of the conditional distribution F(.|x) of Y given X=x suffers from inconsistency problems in the right tail (Beran, 1981), and hence the location function m(x) cannot be estimated consistently in a completely nonparametric way, whenever m(x) involves the right tail of F(.|x) (like e.g. for the conditional mean). We propose two alternative estimators of m(x), that do not share the above inconsistency problems. The idea is to make use of the assumed location-scale model, in order to improve the estimation of F(.|x), especially in the right tail. We obtain the asymptotic properties of the two proposed estimators of m(x). Simulations show that the proposed estimators outperform the completely nonparametric estimator in many cases.

Inductors are extensively used in the design of radio-frequency circuits. In the last decade, the integration of passive components, especially inductors on silicon chips, has led to the widespread development and implementation of Radio Frequency Integrated Circuits (RFICs) in CMOS technologies. However, on-chip passive inductors occupy a large silicon chip area and hardly scale down with technology scaling. Therefore, on-chip passive inductors become formidable obstacles to the realization of highly dense RFICs to be integrated with other highly dense digital circuits on a single chip using a common fabrication process. In recent years, researchers have focused on replacing passive inductors with transistor-only active circuits, namely active inductors. Active inductors can be realized with only a few transistors, which scale down with technology scaling. Therefore, they occupy a fraction of the chip area of their passive counterparts, and can be implemented densely in CMOS processes. Unlike passive inductors, bias dependent operations of active inductors allow for the tuning of their inductance and quality factor Q, and in turn, tuning the performance parameters of RFICs. This thesis focuses on the design and development of passive inductorless CMOS RFICs for ultra-wideband (UWB) receiver front-ends using active inductors. A new Q-enhanced and a new bandwidth-extended tunable active inductors are designed. Using the Q-enhanced active inductor, two tunable UWB low-noise amplifiers (LNAs) (two-stage and three-stage UWB LNAs), a UWB mixer and a wideband local-oscillator (LO) driver are designed. Active inductors are utilized to develop a novel wideband active shunt-peaking technique that decreases high-frequency losses to yield a flat gain over a wide bandwidth. A tunable multiband-UWB front-end integrating a two-stage UWB LNA, and a pair of UWB mixers driven by a pair of wideband LO drivers, is fabricated in a 90nm digital CMOS process. The passive inductorless two-stage UWB LNA, three-stage UWB LNA and UWB front-end occupy chip areas of only 0.0114mm2, 0.0227mm2, and 0.1485mm2, respectively. The active CMOS UWB front-end exhibits a measured flat gain of 22.5dB over 2.5-8.8 GHz bandwidth, and its tunability allows for varying the gain and bandwidth. / Integrated Circuits and Systems

The last few decades have been a very exciting period in thedevelopment of micro-electronics and brought us to the brink ofimplementing entire systems on a single chip, on a hithertounimagined scale. However an unforeseen challenge has croppedup in the form of managing wires, which have become the mainbottleneck in performance, masking the blinding speed of activedevices. A major problem is that increasingly complicatedeffects need to be modelled, but the computational complexityof any proposed model needs to be low enough to allow manyiterations in a design cycle. This thesis addresses the issue of closed form modelling ofthe response of coupled interconnect systems. Following astrict mathematical approach, second order models for thetransfer functions of coupled RC trees based on the first andsecond moments of the impulse response are developed. The2-pole-1-zero transfer function that is the best possible fromthe available information is obtained for the signal path fromeach driver to the output in multiple aggressor systems. Thisallows the complete response to be estimated accurately bysumming up the individual waveforms. The model represents theminimum complexity for a 2-pole-1-zero estimate, for this classof circuits. Also proposed are new techniques for the optimisation ofwires in on-chip buses. Rather than minimising the delay overeach individual wire, the configuration that maximises thetotal bandwidth over a number of parallel wires isinvestigated. It is shown from simulations that there is aunique optimal solution which does not necessarily translate tothe maximum possible number of wires, and in fact deviatesconsiderably from it when the resources available for repeatersare limited. Analytic guidelines dependent only on processparameters are derived for optimal sizing of wires andrepeaters. Finally regular tiled architectures with a commoncommunication backplane are being proposed as being the mostefficient way to implement systems-on-chip in the deepsubmicron regime. This thesis also considers the feasibility ofimplementing a regular packet-switched network-on-chip in atypical future deep submicron technology. All major physicalissues and challenges are discussed for two differentarchitectures and important limitations are identified.

Due to the tremendous growth of traffic volume caused by both exponential increase of number of Internet users and continual emergence of new bandwidth demanding applications, high capacity networks are required in order to satisfactorily handle the extremely large amount of traffic. Hence, optical fiber communication is the key technology for the network infrastructure. This thesis addresses design, analysis and simulation of access and core networks targeting important research problems, which need to be tackled for the effective realization of next generation optical networks. Among different fiber access architectures, passive optical network (PON) is considered as the most promising alternative for the last mile connection due to its relatively low cost and resource efficiency. The inherent bursty nature of the user generated traffic results in dynamically changing bandwidth demand on per subscriber basis. In addition, access networks are required to support differentiated quality of service and accommodate multiple service providers. To address these problems we proposed three novel scheduling algorithms to efficiently realize dynamic bandwidth allocation in PON, along with guaranteeing both the priority and fairness of the differentiated services among multiple users and/or service providers. Meanwhile, because of the increasing significance of reliable access to network services, an efficient fault management mechanism needs to be provided in PON. In addition, access networks are very cost sensitive and the cost of protection should be kept as low as possible. Therefore, we proposed three novel cost-effective protection architectures keeping in mind that reliability requirement in access networks should be satisfied at the minimal cost. Regarding the optical core networks, replacing electronic routers with all-optical switching nodes can offer significant advantages in realizing high capacity networks. Because of the technological limitations for realizing all-optical nodes, the focus is put on the ingenious architecture design. Therefore, we contributed on novel switching node architectures for optical circuit and packet switching networks. Furthermore, we addressed different aspects of routing and wavelength assignment (RWA) problem, which is an important and hard task to be solved in wavelength routed networks. First, we proposed an approach based on the information summary protocol to reduce the large amount of control overhead needed for dissemination of the link state information in the case of adaptive routing. In addition, transparency in optical networks may cause vulnerability to physical layer attacks. To target this critical security related issue, we proposed an RWA solution to minimize the possible reachability of a jamming attack. Finally, in order to evaluate our ideas we developed two tailor-made simulators based on discrete event driven system for the detailed studies of PON and switched optical networks. Moreover, the proposed tabu search heuristic for our RWA solution was implemented in C++. / QC 20100707

Home networks are evolving vigorously and are also becoming more and more heterogeneous and complex. Home networks contain new services and numerous devices with wired and wireless links. At the same time, users demand high levels of quality of service for many new applications. Hence, the supervision of links quality tends to be mandatory in emergent home networks, to trigger QoS mechanisms. In this context, this dissertation proposes the utilization of the available bandwidth as a performance indicator in hybrid home networks (e.g. for remote service-evaluation, diagnosis and fault detection) and as link-state metric for a number of QoS mechanisms (e.g., admission control, path selection and load balancing). In this dissertation, we firstly explain why available bandwidth probing is a fundamental tool in the new QoS architectures for hybrid home networks and we explore different use cases. We investigate the networking constraints that affect bandwidth probing in hybrid home networks. We explain why most tools to measure available bandwidth on Internet paths are not very efficient on home networks. We make a taxonomic study of a number of state-of-the-art probing techniques and probing tools for available bandwidth. At that aim, we propose a functional framework, called Metrics Profile Capture Filter and Feedback (MPCFF). Based on these studies, we choose Iperf in TCP mode as an attractive tool to estimate available bandwidth. Then, we conduct a performance evaluation on an experimental test bed, to compare Iperf in lightweight TCP-mode vs. various state-of-the-art tools (Wbest, Pathchirp, Pathload and IGI/PTR). We show that Iperf in lightweight TCP mode is superior in terms of accuracy and speed of convergence. Finally, we integrate Iperf in lightweight TCP mode into a path selection protocol using the Inter-MAC software (developed in the Omega European project). We show, on a test bed, how Iperf in lightweight TCP mode enhances the Inter-MAC path selection by eliminating performance deterioration due to time-varying capacity links. Our results show that Iperf in lightweight TCP mode, accurately, triggers path selection and load balancing mechanisms, to adapt the network resources to the strong degradation of WiFi and PLC links. Available bandwidth probing detects and helps to control real-time deviations of the quality of experience, associated of sensitive applications. At the same time, available-bandwidth probing helps the service provider to isolate failures and to monitor the end-to-end residential service, on demand or on long-term periodicity basis.

Scalable on-demand content delivery systems, designed to effectively handle increasing request rates, typically use service aggregation or content replication techniques. Service aggregation relies on one-to-many communication techniques, such as multicast, to efficiently deliver content from a single sender to multiple receivers. With replication, multiple geographically distributed replicas of the service or content share the load of processing client requests and enable delivery from a nearby server.<p>Previous scalable protocols for downloading large, popular files from a single server include batching and cyclic multicast. Analytic lower bounds developed in this thesis show that neither of these protocols consistently yields performance close to optimal. New hybrid protocols are proposed that achieve within 20% of the optimal delay in homogeneous systems, as well as within 25% of the optimal maximum client delay in all heterogeneous scenarios considered.<p>In systems utilizing both service aggregation and replication, well-designed policies determining which replica serves each request must balance the objectives of achieving high locality of service, and high efficiency of service aggregation. By comparing classes of policies, using both analysis and simulations, this thesis shows that there are significant performance advantages in using current system state information (rather than only proximities and average loads) and in deferring selection decisions when possible. Most of these performance gains can be achieved using only local (rather than global) request information.<p>Finally, this thesis proposes adaptations of already proposed peer-assisted download techniques to support a streaming (rather than download) service, enabling playback to begin well before the entire media file is received. These protocols split each file into pieces, which can be downloaded from multiple sources, including other clients downloading the same file. Using simulations, a candidate protocol is presented and evaluated. The protocol includes both a piece selection technique that effectively mediates the conflict between achieving high piece diversity and the in-order requirements of media file playback, as well as a simple on-line rule for deciding when playback can safely commence.

Channel coding has been largely used for the purpose of improving error performance on a communications system. Typical methods based on added redundancy allow for error detection and correction, this improvement however comes at a cost of bandwidth. This thesis focuses on channel coding for the bandwidth-limited channel where no bandwidth expansion is allowed. We first discuss the idea of coding for the bandwidth-limited channel as seen from the signal space point of view where the purpose of coding is to maximize the Euclidian distance between constellation points without increasing the total signal power and under the condition that no extra bits can be added. We then see the problem from another angle and identify the tradeoffs related to bandwidth and error performance. This thesis intends to find a simple way of achieving an improvement in error performance for the bandwidth-limited channel without the use of lattice codes or trellis-coded modulation. The proposed system is based on convolutional coding followed by multilevel transmission. It achieved a coding gain of 2 dB on Eb/No or equivalently, a coding gain of approximately 2.7 dB on SNRnorm without increase in bandwidth. This coding gain is better than that obtained by a more sophisticated lattice code Gosset E8 at the same error rate.

Scalable on-demand content delivery systems, designed to effectively handle increasing request rates, typically use service aggregation or content replication techniques. Service aggregation relies on one-to-many communication techniques, such as multicast, to efficiently deliver content from a single sender to multiple receivers. With replication, multiple geographically distributed replicas of the service or content share the load of processing client requests and enable delivery from a nearby server.<p>Previous scalable protocols for downloading large, popular files from a single server include batching and cyclic multicast. Analytic lower bounds developed in this thesis show that neither of these protocols consistently yields performance close to optimal. New hybrid protocols are proposed that achieve within 20% of the optimal delay in homogeneous systems, as well as within 25% of the optimal maximum client delay in all heterogeneous scenarios considered.<p>In systems utilizing both service aggregation and replication, well-designed policies determining which replica serves each request must balance the objectives of achieving high locality of service, and high efficiency of service aggregation. By comparing classes of policies, using both analysis and simulations, this thesis shows that there are significant performance advantages in using current system state information (rather than only proximities and average loads) and in deferring selection decisions when possible. Most of these performance gains can be achieved using only local (rather than global) request information.<p>Finally, this thesis proposes adaptations of already proposed peer-assisted download techniques to support a streaming (rather than download) service, enabling playback to begin well before the entire media file is received. These protocols split each file into pieces, which can be downloaded from multiple sources, including other clients downloading the same file. Using simulations, a candidate protocol is presented and evaluated. The protocol includes both a piece selection technique that effectively mediates the conflict between achieving high piece diversity and the in-order requirements of media file playback, as well as a simple on-line rule for deciding when playback can safely commence.

As demand for higher bandwidth I/O grows, the front end design of serial link becomes significant to overcome stringent timing requirements on noisy and bandwidthlimited channels. As a clock reconstructing module in a receiver, the recovered clock quality of Clock and Data Recovery is the main issue of the receiver performance. However, from unknown incoming jitter, it is difficult to optimize loop dynamics to minimize steady-state and dynamic jitter. In this thesis a 10 Gb/s adaptive loop bandwidth clock and data recovery circuit with on-chip loop filter is presented. The proposed system optimizes the loop bandwidth adaptively to minimize jitter so that it leads to an improved jitter tolerance performance. This architecture tunes the loop bandwidth by a factor of eight based on the phase information of incoming data. The resulting architecture performs as good as a maximum fixed loop bandwidth CDR while tracking high speed input jitter and as good as a minimum fixed bandwidth CDR while suppressing wide bandwidth steady-state jitter. By employing a mixed mode predictor, high updating rate loop bandwidth adaptation is achieved with low power consumption. Another relevant feature is that it integrates a typically large off-chip filter using a capacitance multiplication technique that employs dual charge pumps. The functionality of the proposed architecture has been verified through schematic and behavioral model simulations. In the simulation, the performance of jitter tolerance is confirmed that the proposed solution provides improved results and robustness to the variation of jitter profile. Its applicability to industrial standards is also verified by the jitter tolerance passing SONET OC-192 successfully.